INDEX

Explanations

words related to misinformation or misrepresentation

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

starting with "mis" (meaning wrong/bad)

np_acts-logits-general · gemini-2.0-flash

misplaced terms and negative concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

oredCriteria

-0.58

ClientSize

-0.53

InitVars

-0.52

età

-0.52

 BIBSYS

-0.52

 trampa

-0.51

Abit

-0.50

 ferma

-0.48

 nawr

-0.48

POSITIVE LOGITS

 interpr

0.57

representation

0.57

ogyn

0.55

haps

0.55

matched

0.54

 XCTest

0.54

interpretation

0.54

interpret

0.53

ogy

0.51

fits

0.50

Activations Density 0.162%

words related to misinformation or misrepresentation

starting with "mis" (meaning wrong/bad)

misplaced terms and negative concepts

No Comments

No Known Activations