INDEX

Explanations

mentions of specific names or entities related to Duplicate Entry

oai_token-act-pair · gpt-3.5-turbo

references to specific place names and named entities

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

rawdownloadcloneembedreportprint

-0.81

 compr

-0.79

tics

-0.78

ulative

-0.78

ESA

-0.71

iannopoulos

-0.71

stretched

-0.70

channelAvailability

-0.69

nces

-0.69

BILITY

-0.68

POSITIVE LOGITS

 Cities

1.05

 Falls

1.03

 Peaks

1.03

ergy

0.92

hower

0.88

ĸļ

0.84

 Lakes

0.80

 Timbers

0.79

 Rapids

0.78

erton

0.78

Activations Density 0.028%

mentions of specific names or entities related to Duplicate Entry

references to specific place names and named entities

No Comments

No Known Activations