INDEX

Explanations

names of people or entities

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names and terms associated with various people and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 compan

-0.75

natureconservancy

-0.70

adena

-0.69

 flares

-0.67

akeru

-0.66

 showc

-0.64

achine

-0.64

 propos

-0.62

 Respond

-0.62

 outp

-0.61

POSITIVE LOGITS

bart

0.87

 Claw

0.77

hyde

0.72

 Choi

0.72

 Syndrome

0.69

meyer

0.67

lings

0.66

 Elliott

0.66

ulty

0.66

bush

0.66

Activations Density 0.230%

names of people or entities

mentions of specific names and terms associated with various people and events

No Comments

No Known Activations