INDEX

Explanations

the word "common" in various contexts, relating to shared ideas, values, or characteristics

oai_token-act-pair · gpt-3.5-turbo

references to shared or collective concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

enda

-0.74

asus

-0.72

agate

-0.72

ocalypse

-0.72

resy

-0.71

gur

-0.70

oby

-0.69

hyde

-0.69

otos

-0.69

gy

-0.68

POSITIVE LOGITS

wealth

1.28

alities

0.98

 ancestor

0.93

ensical

0.90

 denomin

0.86

ality

0.85

places

0.84

 occurrence

0.76

 Common

0.76

 common

0.74

Activations Density 0.016%

the word "common" in various contexts, relating to shared ideas, values, or characteristics

references to shared or collective concepts

No Comments

No Known Activations