INDEX

Explanations

character strings that seem to be unrelated to the natural language text

oai_token-act-pair · gpt-3.5-turbo

numerical data and identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

soDeliveryDate

-0.77

 reluct

-0.76

Spread

-0.75

Tracker

-0.73

Repe

-0.70

 Sisters

-0.69

soType

-0.68

 advoc

-0.68

âĸ¬

-0.68

sein

-0.68

POSITIVE LOGITS

ffe

1.05

fe

1.00

cf

0.98

afa

0.94

efe

0.91

fc

0.90

0.89

0.88

Activations Density 0.054%

character strings that seem to be unrelated to the natural language text

numerical data and identifiers

No Comments

No Known Activations