INDEX

Explanations

phrases indicating opposition or disagreement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

jména

-0.73

 houſe

-0.69

nemo

-0.67

ERTY

-0.66

 florales

-0.65

Txn

-0.65

談社

-0.64

 Wren

-0.64

 placés

-0.61

 собі

-0.60

POSITIVE LOGITS

 Against

2.03

Against

2.02

 against

1.91

against

1.84

 AGAINST

1.80

 gegen

1.48

 contre

1.35

 tegen

1.24

 melawan

1.14

 против

1.12

Activations Density 0.062%

phrases indicating opposition or disagreement

No Comments

No Known Activations