INDEX

Explanations

proper nouns and names

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

attributing speech to names

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 transfieras

-0.70

 betweenstory

-0.69

OGND

-0.68

EndContext

-0.68

+#+#

-0.66

ंदीखरीदारी

-0.64

AISSEE

-0.62

 nonUne

-0.62

uxxxx

-0.61

 oprot

-0.60

POSITIVE LOGITS

 referring

0.75

 refiri

0.66

 quoted

0.65

 dikutip

0.62

 about

0.62

 tungkol

0.60

 quoting

0.60

 speaking

0.57

 regarding

0.57

 explaining

0.57

Activations Density 0.017%

proper nouns and names

attributing speech to names

No Comments

No Known Activations

proper nouns and names

attributing speech to names

No Comments

No Known Activations