INDEX

Explanations

instances of the word "Last" followed by a number indicating a past event or timeframe

oai_token-act-pair · gpt-4o-mini Triggered by @bot

last [time unit] references

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

XtraReports

-0.46

MNA

-0.41

 jsPsych

-0.41

<eos>

-0.38

 OMITBAD

-0.38

RUnlock

-0.36

-0.35

↵↵

-0.35

CFA

-0.35

CPF

-0.34

POSITIVE LOGITS

majánló

0.71

SequentialGroup

0.61

findpost

0.59

 zoude

0.59

ſcher

0.58

 Wikiseite

0.58

umably

0.57

ðsíða

0.55

cemment

0.54

hyrchwyd

0.54

Activations Density 0.006%

instances of the word "Last" followed by a number indicating a past event or timeframe

last [time unit] references

No Comments

No Known Activations

instances of the word "Last" followed by a number indicating a past event or timeframe

last [time unit] references

No Comments

No Known Activations