INDEX

Explanations

terms related to stopping, slowing, and hesitating actions

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

">//

-0.73

 Waugh

-0.69

Бахар

-0.69

GLenum

-0.68

."));

-0.68

 كومونز

-0.66

.")]

-0.66

ftagPool

-0.65

CppCodeGen

-0.64

ⓧ

-0.64

POSITIVE LOGITS

 stop

1.18

 stopped

1.15

 pause

1.15

 halt

1.09

 STOP

1.07

 halted

1.07

 Stop

1.06

 stops

1.04

stopped

1.02

STOP

1.01

Activations Density 0.355%

terms related to stopping, slowing, and hesitating actions

No Comments

No Known Activations