INDEX

Explanations

the word "all"

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases indicating inclusivity or universality

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

abwe

-0.70

 nowhere

-0.69

 tremend

-0.67

rir

-0.66

SHIP

-0.66

 Kamp

-0.62

chie

-0.61

flation

-0.61

hran

-0.60

 Fidel

-0.58

POSITIVE LOGITS

ocating

1.14

igator

1.12

 kinds

1.11

iances

1.06

 sorts

1.02

igators

0.98

iance

0.97

ocation

0.95

uv

0.94

usions

0.94

Activations Density 0.124%

the word "all"

repetitive phrases indicating inclusivity or universality

No Comments

No Known Activations