INDEX

Explanations

technical terms and code structure references

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_131k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.33.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 found

-0.90

 Found

-0.82

found

-0.80

 findings

-0.73

Found

-0.70

 finding

-0.70

 FOUND

-0.66

 finds

-0.62

FOUND

-0.62

Findings

-0.60

POSITIVE LOGITS

AndEndTag

0.56

ScopeManager

0.49

 nakalista

0.47

EndTag

0.46

 solace

0.45

 coupable

0.44

isContained

0.43

 استنادى

0.43

 fault

0.43

帖最后由

0.43

Activations Density 0.034%

technical terms and code structure references

No Comments

No Known Activations