INDEX

Explanations

phrases containing the word "uns" followed by another term, indicating a negative context

oai_token-act-pair · gpt-3.5-turbo

occurrences of the prefix "uns" in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 briefs

-0.75

tsky

-0.74

phrine

-0.73

SHIP

-0.73

 Rams

-0.72

 Dynamics

-0.72

 Nanto

-0.69

 Blazers

-0.68

 Guardians

-0.67

 stages

-0.67

POSITIVE LOGITS

rep

1.08

aved

1.08

avour

1.08

killed

1.05

heat

1.01

olicited

1.01

oci

1.00

iders

1.00

apon

1.00

ided

0.97

Activations Density 0.012%

phrases containing the word "uns" followed by another term, indicating a negative context

occurrences of the prefix "uns" in words

No Comments

No Known Activations