INDEX

Explanations

phrases related to reported occurrences or actions

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "reportedly" indicating sources of information or rumors

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

allas

-0.68

lap

-0.67

arium

-0.66

uit

-0.65

eem

-0.65

cest

-0.64

 Frog

-0.63

ĸļ

-0.63

gers

-0.63

Geh

-0.63

POSITIVE LOGITS

 obtained

0.75

 located

0.74

 priced

0.74

 complied

0.70

Ĥİ

0.69

 solic

0.69

icably

0.68

 exercised

0.67

 entertained

0.67

 headed

0.67

Activations Density 0.020%

phrases related to reported occurrences or actions

mentions of the word "reportedly" indicating sources of information or rumors

No Comments

No Known Activations