INDEX

Explanations

words related to making adjustments or changes

oai_token-act-pair · gpt-3.5-turbo

terms related to adjustment or modification processes

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

gom

-0.73

OGR

-0.72

################################

-0.69

 Furious

-0.68

DoS

-0.68

advertisement

-0.68

Write

-0.67

INST

-0.66

Å

-0.66

OPA

-0.66

POSITIVE LOGITS

 adjustments

0.98

 adjustment

0.98

 adjust

0.95

 adjusting

0.89

ments

0.85

adjust

0.84

 adjusts

0.84

 straps

0.79

 adjusted

0.77

 Adjust

0.77

Activations Density 0.016%

words related to making adjustments or changes

terms related to adjustment or modification processes

No Comments

No Known Activations