INDEX

Explanations

code-related terms and structures, particularly within programming and configuration contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

 ſte

-0.69

 ſta

-0.69

 houſe

-0.63

ſt

-0.61

 faſt

-0.60

 ſch

-0.59

 purpoſe

-0.59

 leaſt

-0.58

 leſs

-0.57

 ſtre

-0.57

POSITIVE LOGITS

back

0.77

auto

0.77

pro

0.77

all

0.75

info

0.74

list

0.73

string

0.73

per

0.73

key

0.73

time

0.73

Activations Density 2.026%

code-related terms and structures, particularly within programming and configuration contexts

No Comments

No Known Activations