INDEX

Explanations

references to various types of fees, highlighting a focus on high or significant fees

oai_token-act-pair · gpt-3.5-turbo

references to financial charges or fees

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãĥŁ

-0.71

 Flesh

-0.69

search

-0.68

uve

-0.68

 Dwell

-0.67

 Antarctic

-0.67

 Remem

-0.66

pse

-0.66

 Surface

-0.66

void

-0.65

POSITIVE LOGITS

 fees

1.51

 Fees

1.14

fee

1.08

 waivers

1.00

heet

0.97

 payable

0.87

 incurred

0.87

 levied

0.86

umed

0.85

 dues

0.85

Activations Density 0.009%

references to various types of fees, highlighting a focus on high or significant fees

references to financial charges or fees

No Comments

No Known Activations