INDEX

Explanations

occurrences of the word "try" and its variations in context relating to attempts or actions

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

try followed by decoder or visitor

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

 betweenstory

-0.75

ſelf

-0.59

ſelves

-0.54

 ſche

-0.54

 myſelf

-0.53

 itſelf

-0.50

dafx

-0.50

 themſelves

-0.49

 ſou

-0.49

 pleaſure

-0.47

POSITIVE LOGITS

try

0.60

Try

0.58

try

0.58

DockStyle

0.57

setopt

0.57

Try

0.53

TRY

0.52

requireNonNull

0.52

 trypsin

0.46

 stepwise

0.46

Activations Density 0.005%

occurrences of the word "try" and its variations in context relating to attempts or actions

try followed by decoder or visitor

No Comments

No Known Activations