INDEX

Explanations

phrases that express negation or limitation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 itſelf

-1.12

 raiſ

-0.96

 himſelf

-0.95

 myſelf

-0.90

ſelf

-0.88

 ſche

-0.86

 pleaſure

-0.86

 ſch

-0.86

 cauſe

-0.85

 ſta

-0.84

POSITIVE LOGITS

 schon

1.05

 noch

1.00

nog

0.86

 nicht

0.85

 auch

0.74

nicht

0.72

 niet

0.71

nur

0.69

 immer

0.68

Noch

0.68

Activations Density 0.139%

phrases that express negation or limitation

No Comments

No Known Activations