INDEX

Explanations

linguistic structures that indicate interactions and presentations

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

 guerr

-0.30

 Erweiterung

-0.29

 ampliación

-0.28

 incontr

-0.28

 rifer

-0.28

 Erscheinung

-0.28

 inmediato

-0.27

 förr

-0.27

 assolu

-0.27

 réserve

-0.27

POSITIVE LOGITS

DockStyle

0.71

 ब्रेकडाउन

0.69

 Infórmanos

0.66

addGap

0.64

'{@

0.63

writeField

0.63

 nakalista

0.63

ftagPool

0.61

 overpriced

0.60

bootstrapcdn

0.60

Activations Density 0.087%

linguistic structures that indicate interactions and presentations

No Comments

No Known Activations