INDEX

Explanations

references to presence and observation in various contexts

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 työn

-0.49

 Chwiliwch

-0.48

 desmotivaciones

-0.47

 fieltro

-0.47

Organisateur

-0.46

 naudoti

-0.46

 femeninos

-0.46

 sauvages

-0.46

 variedade

-0.45

 trabajos

-0.45

POSITIVE LOGITS

着

0.44

 cameras

0.44

BASELINE

0.43

 CCTV

0.43

giphy

0.43

watched

0.42

vertre

0.42

natal

0.41

nessed

0.41

 eyes

0.40

Activations Density 0.031%

references to presence and observation in various contexts

No Comments

No Known Activations