INDEX

Explanations

identifiers or references related to messages

oai_token-act-pair · gpt-3.5-turbo

references to messages 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

engeance

-0.82

ONSORED

-0.80

aughs

-0.79

rowd

-0.77

cling

-0.73

ESCO

-0.71

itals

-0.71

biased

-0.70

orks

-0.69

nec

-0.69

POSITIVE LOGITS

 messages

1.09

 Messages

0.99

 goodbye

0.93

Message

0.93

 message

0.93

message

0.91

FontSize

0.87

mith

0.84

achine

0.81

 inbox

0.81

Activations Density 0.019%

identifiers or references related to messages

references to messages in various contexts

No Comments

No Known Activations