INDEX

Explanations

references to specific concepts or names

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "iq."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

andon

-0.73

 Victory

-0.70

 BASE

-0.68

 pitching

-0.65

 mell

-0.65

Ach

-0.63

 yard

-0.63

grounds

-0.63

 fungus

-0.61

 starters

-0.60

POSITIVE LOGITS

iq

4.89

IQ

1.66

qa

1.59

eq

1.13

iques

1.09

qq

1.03

QL

1.01

1.00

MQ

0.96

ique

0.96

Activations Density 0.027%

references to specific concepts or names

occurrences of the term "iq."

No Comments

No Known Activations