INDEX

Explanations

the word "then" in sequences

oai_token-act-pair · gpt-3.5-turbo

the phrase "but then."

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

toe

-0.77

undo

-0.70

orde

-0.69

ction

-0.67

ters

-0.67

idad

-0.65

chy

-0.65

illy

-0.63

OUR

-0.62

ours

-0.61

POSITIVE LOGITS

 conclud

0.77

anski

0.72

Ļ½

0.71

 suddenly

0.69

©¶æ

0.69

 proceeded

0.68

 abruptly

0.68

ulhu

0.67

etheless

0.66

srf

0.64

Activations Density 0.046%

the word "then" in sequences

the phrase "but then."

No Comments

No Known Activations