INDEX

Explanations

references to specific cultural or geographical terms, particularly "Tonga"

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Tong" in various contexts

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

ãĥ¯ãĥ³

-0.90

 Elys

-0.82

 Stores

-0.79

âķĲâķĲ

-0.77

 GOODMAN

-0.74

 Aether

-0.73

 Trance

-0.69

puter

-0.68

ICAN

-0.67

IST

-0.66

POSITIVE LOGITS

ues

1.16

 Tong

1.12

 tong

0.90

ueless

0.87

ue

0.86

raph

0.86

ued

0.84

atu

0.83

uing

0.79

ug

0.79

Activations Density 0.026%

references to specific cultural or geographical terms, particularly "Tonga"

mentions of the word "Tong" in various contexts

No Comments

No Known Activations