INDEX

Explanations

references to connections between characters in a story

oai_token-act-pair · gpt-3.5-turbo

the conjunction "and" in various contexts, indicating a focus on connection or addition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ÑĮ

-0.73

rued

-0.73

anmar

-0.70

atars

-0.70

Ñı

-0.69

Were

-0.68

onica

-0.65

auga

-0.65

igate

-0.65

oward

-0.65

POSITIVE LOGITS

 prefers

1.73

 enjoys

1.69

 understands

1.61

 knows

1.60

 wants

1.60

 believes

1.58

 spends

1.56

 loves

1.55

 intends

1.54

 thinks

1.53

Activations Density 0.329%

references to connections between characters in a story

the conjunction "and" in various contexts, indicating a focus on connection or addition

No Comments

No Known Activations