INDEX

Explanations

words related to sparking or causing events or reactions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "sparked" in relation to events or reactions

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

esan

-0.91

uters

-0.78

itte

-0.76

ibble

-0.68

endi

-0.67

 wives

-0.65

 Neighbor

-0.64

 grade

-0.58

azel

-0.58

opers

-0.58

POSITIVE LOGITS

TextColor

0.97

 sparked

0.94

rings

0.89

 sparking

0.87

actionDate

0.81

 ignited

0.77

 stir

0.77

 ignite

0.74

lighting

0.73

lights

0.73

Activations Density 0.020%

words related to sparking or causing events or reactions

occurrences of the word "sparked" in relation to events or reactions

No Comments

No Known Activations