INDEX

Explanations

phrases related to comparisons and levels

oai_token-act-pair · gpt-3.5-turbo

adjectives and descriptors indicating scale or intensity

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

 Learns

-0.63

ONEY

-0.59

KEY

-0.57

 Wilde

-0.55

 Kills

-0.55

 Trials

-0.55

 Officers

-0.54

 Hernandez

-0.53

Spa

-0.53

 Auth

-0.52

POSITIVE LOGITS

tenance

1.20

terday

1.04

lihood

1.00

etheless

0.84

ãĤ´ãĥ³

0.79

theless

0.76

usterity

0.75

/-

0.74

itarian

0.73

itored

0.73

Activations Density 0.232%

phrases related to comparisons and levels

adjectives and descriptors indicating scale or intensity

No Comments

No Known Activations