INDEX

Explanations

phrases indicating a general truth or statement

oai_token-act-pair · gpt-3.5-turbo

statements indicating similarity or parallels between different subjects or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 channels

-0.62

bro

-0.61

 contrace

-0.60

 discovery

-0.60

dun

-0.59

 diarr

-0.58

vironment

-0.58

 diversion

-0.57

 interrupts

-0.56

 bang

-0.55

POSITIVE LOGITS

IELD

0.80

 apply

0.79

VB

0.77

 applies

0.75

blance

0.71

ï¸

0.71

pez

0.70

":"/

0.70

etry

0.69

amiya

0.69

Activations Density 0.245%

phrases indicating a general truth or statement

statements indicating similarity or parallels between different subjects or concepts

No Comments

No Known Activations