INDEX

Explanations

phrases related to significant events or decisions

oai_token-act-pair · gpt-3.5-turbo

quantifiable expressions of quantity or extent

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

 laun

-0.59

iru

-0.54

inki

-0.52

AP

-0.51

irit

-0.51

Buff

-0.51

 Cats

-0.50

 remembers

-0.49

article

-0.49

reat

-0.49

POSITIVE LOGITS

 lately

1.05

 since

0.89

 unsuccessfully

0.72

 numerous

0.70

 recently

0.68

since

0.68

 countless

0.67

successfully

0.65

 administrations

0.61

 innumerable

0.58

Activations Density 1.122%

phrases related to significant events or decisions

quantifiable expressions of quantity or extent

No Comments

No Known Activations