INDEX

Explanations

words related to completing actions or achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

supset

-0.76

bytes

-0.75

ROLE

-0.70

triangleq

-0.66

pyr

-0.66

brigens

-0.66

 Pyle

-0.66

 Krä

-0.65

 Vallarta

-0.62

 hearsay

-0.62

POSITIVE LOGITS

 finish

2.11

 FINISH

2.04

 Finish

2.03

 finishes

1.92

 finished

1.91

finish

1.91

Finish

1.84

 finishing

1.83

FINISH

1.80

 Finished

1.77

Activations Density 0.036%

words related to completing actions or achievements

No Comments

No Known Activations