INDEX

Explanations

mentions of transcripts in various contexts

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

transcript 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

rungsseite

-0.44

 Italijani

-0.37

Portail

-0.37

AddTagHelper

-0.36

errHandler

-0.35

تقاوى

-0.35

AndEndTag

-0.34

贛

-0.33

enumi

-0.33

 Савезне

-0.32

POSITIVE LOGITS

 Transcript

0.84

Transcript

0.66

 transcript

0.66

setVerticalGroup

0.60

transcript

0.60

 transcripts

0.59

 transcribed

0.56

исленность

0.55

CodedInputStream

0.53

RepeatedField

0.50

Activations Density 0.000%

mentions of transcripts in various contexts

transcript references

No Comments

No Known Activations